Framework for Generalization and Improvement of Relational Data Warehouses

Abstract

In this paper we present the system of relational data warehouse optimization. The system is based on generalized definition of solution space which includes lot of factors relevant for performance of the relational data warehouses. Novelties are hybrid Greedy - Genetic algorithms applied to generalized solution space and specific techniques for improvement of the efficiency of optimization process. Experimental results shows that our algorithms outperforms traditional Genetic algorithms in the speed of finding optimal (or near optimal) solutions.